Classic controller DEFCON 4: FET Drive Error

[Dmitry]

I'll say it again.
On the blue terminal is already present difference 6 volt, from indications on the display of the controller.
The voltage on a blue terminals equally with the battery voltage.

And, as I have defined this difference is stored in different modes charging.
Note that after upgrading the software, all controller settings were reset to default. No correction voltage is not installed. Correction battery voltage temperature is turned off.

Mode "BULK MPPT " , such as the controller shows 51,0 volt battery life measured by the control voltage tester will be at 6 volts more 57,0 volts.
Mode "ABSORB" the controller shows 54,8 volt battery control tester voltage measured 60,8 Volta, in the settings 57,2 volts.
Mode "FLOAT" the controller shows 52,7 volt battery control tester voltage measured 58,7 Volta, in the settings 54,4 volts.

Write your thoughts, today, tomorrow can do with the controller any manipulation, to give access to a laptop with Midnite Application via TeamViewer.
Then go away and most likely will turn off, as it is dangerous to leave such unpredictable controller without attention.

[Dmitry]

Colleagues managed to catch spontaneous reboot the controller.
It looks like this.
The controller is charging. The panel displays all the settings. Unexpectedly inside the body begins flashing red led. On the Board it is marked D2. After 3 seconds it goes off on the display of the controller at this time, see termination of charging, the battery voltage and solar panels are displayed. Further, this also led (D2) starts poorly lit, run the fan, blue led is lit and the controller is reset, and the panel does not go out, but still displays the voltage of the battery and solar panel. After 1 minute charging begins on the scoreboard displays all parameters.
When you restart also loses communication with the controller application Midnite Application.


Waiting for your considerations.
If the answer please do not mix the problem controller restarts and incorrect voltage, separate posts for better understanding of the problem.

[boB]

Dmitry, when you see the Classic reset (flashing D2, etcetera), what is the MNGP/Remote LCD on the Classic displaying at that time ?

Is the remote MNGP displaying logging information ?   I ask this because if the Local App panel and the MNGP/remote LCD is
both displaying the logging information at the same time, this may in some cases make the Classic reset itself.

If that is not the case, then you may just have a defective charge controller.  I will let Ryan (Half Crazy)  respond as well.

Sorry about these problems.

[Dmitry]

boB I can't exactly answer your question, because of the chaotic controller restarts.
Unfortunately, at the time that I got to see I didn't bring a camera to take pictures of the process.
Reboot very difficult to catch, I tried yesterday to sit and wait, but the controller is unfortunately not been restarted.
I drew attention to two points, but cannot guarantee their complete accuracy. This is my personal opinion, I think, maybe it's not.

First.
The controller restarts when intensively looking for a point MPPT.  That is the day when solar panels generate a lot of power reboots I have not committed. In the evening when the power panels falls to between 100 and 150 watts (I have at the peak of the panel give about 1800 watts) the controller begins to actively seek the maximum power point, at this moment begin chaotic reboot.
If you look at the beginning of this topic, you will see that on the previous firmware in these moments controller went in the error DEFCON 4: FET Drive Error
She does not pop up, simply reboot.

A second.
It seemed that the frequency of reboots depends on the temperature of the internal circuit boards of the controller.
Yesterday I measured the voltage on the blue Board and left the top cover of the controller next. Now the frequency of reboots fell sharply. That got me thinking about the overheating of the components inside the device. While fans the controller does not run, I understand he believes that the temperature is normal for him. I estimated the temperature difference, with the cover removed by 1.5 - 2 degrees below it.
Don't even know you believe me, but simply removing the top cover leads to such effect. And to see it I had to spend two days.

It may be the case in a blue cord, he at the time of the removed cover is taut. Could there be a problem in the reboots from the cable? Does it make sense to change it to Yes to another? With mind cable without damage I did nothing.

Now describe reboot time.
To understand in brackets will approximate the displayed values. At the time between reboots.

The controller is charging. The panel displays all the parameters: voltage solar (103,3 volts), the instantaneous power (60 watt), kilowatt hours (0,1), the battery voltage (49,2 volts), the current battery charge (1.2 amps), the operating mode of the controller (Bulk MPPT).

Unexpectedly inside the body begins flashing red led. On the Board it is marked D2. After 3 seconds it goes off on the display of the controller at that time, see: voltage solar (115,0 volts), the instantaneous power (0 watts), kilowatt hours (0), the battery voltage (49,2 volts), the current battery charge (0 amps), the operating mode of the controller (? I can't say).

Further, this also led (D2) starts poorly lit, run the fan, blue led is lit and the controller is reset, and the panel does not go out, but still displays the voltage of the battery and solar panel.

After 1 minute charging begins on the scoreboard displays all parameters kilowatt hours reset to 0.
Voltage of solar panels (103,3 volts), the instantaneous power (60 watt), kilowatt hours (0), the battery voltage (49,2 volts), the current battery charge (1.2 amps), the operating mode of the controller (Bulk MPPT).

When you restart also loses communication with the controller application Midnite Application.

Anything to do with the panel I was not managed then, everything happened so quickly that I only had time to watch and remember.
After the reboot panel is working without problems, all settings are displayed and edited.


boB really bothers me is the problem with stress.
Now I can't leave the controller unattended, he instead to charge the batteries they will simply be roasted increased by 6 volts voltage. What to do I just don't know.
If with the old firmware I was not afraid to leave the controller unattended, with new, it's just dangerous.

You can now record video you if it helps. Unfortunately, I have to leave tomorrow and I won't have direct access to the controller for a while.

Also, I started a post about AUX conclusions
http://midniteforum.com/index.php?topic=1860.0
I thought the firmware of the device will remove the problem, but it is just not up to it. It would have to know exactly what the problem reboots does not depend on AUX conclusions, I was disconnected in all settings turned off, left the controller one. Now he starts without a problem, next to understand until I became.

[boB]

Dmitry, what is the serial number of your Classic ?

Or, the date of manufacture ?

[Dmitry]

Serial 10576.
Production date I do not know where to look.
The checklist quality signed date 9-26-13.

[Dmitry]

Colleagues, boB, post a video link with the increase of the voltage at the battery connected to the controller.
http://youtu.be/SExAEdj1V5Y

I have already thoroughly described in this process, refer to the time of ascending at 6:19.
Also be sure to see, starting from 8:40 to end the recording. Where I make the control voltage measurement.

Important notice.
When writing, for greater clarity of the process I decided to slightly modify the system configuration. Was reduced battery capacity up to 14 ampere hour 48 volts, so you can see a sudden increase in voltage issued by the controller to the battery. If battery capacity is large, this process simply stretched for time-dependent current capacity issued by the solar panels and the charge status of the batteries. But still in the end of the battery trying to enter the increased voltage.

In this video controller only infects battery, no load additional charge to the batteries have not connected, no controller setup at this point had been made, AUX conclusions disabled.

[dgd]

Dmitry,

Your video is out of focus most of the time. When it briefly comes into focus where there is only 0.4v difference between the MNGP and your voltmeter this just shows the TWEAKS menu needs to adjust the display voltage.
After another period of not focussed it comes into focus and we see the voltmeter is suddenlt showing about 6 volts higher than the MNGP.
It would be nice to see the point where this change occured and what the MNGP had displayed. We can see before and after but not the event itself.

56.2v seems to be your absorb voltage setting but meter for no reason changes from 55.8v to 62v. The input voltage stays about the same but watts input has risen from about 45w to 90w. At least the Classic is consistent with battery voltage reading

Have you tested voltages with another voltmeter, one that has been certified as accurate?
WHat guage are those battery cables into the Classic #6?  They look small diameter cable especially if you can get the voltmeter test probes along with the power cables into the blue connector block

When you use the voltmeter to test the actual battery voltage at the battery then what is that voltage?

It also seems extremely unlikely that the battery voltage would change from 55.8 to 62 in a second or so unless some other large charging current was applied by an external device. Makes me think the meter is suspicious.

dgd

[Dmitry]

DGD don't get me wrong, but what is the quality of the video I capture. I believe the quality HD IPhone enough to capture and not all would see. What put the quality on youtube.com worse, it is not dependent on me. However, if you watched this video in full screen and forcibly changed quality with 360p in 720p HD, you would not have written to me that the video is not in focus, all clearly and perfectly visible.
A time of increasing tension I wrote it 6:19.
Look again video as 720p HD.

And now a question for you?
If I have a blue contacts (batteries) terminal blocks of the controller is connected voltmeter and I at the beginning of the video you can see the difference between the voltage of the voltage and the voltage panel controller 0.4 volts, this we believe is inaccurate. Explain to me then. If voltmeter not be disconnected from the contacts on the blue terminal blocks of the controller, which then at 6:19 time video a difference in 6 volts between the readings displayed voltmeter and indications MNGP panel controller?

[dgd]

Dmitry,

OK, maybe its just me but the video is mostly out of focus and a few times it suddenly gets sharp focus.
Maybe someone else can view your video and confirm if it's my problem with viewing in focus.

The 0.4 volt difference is Ok, Maybe just the Classic display voltage needs adjusting using the TWEAKS menu.

The 6 volt difference is just not right.
So why does your voltmeter show 55.8volts then within a second or so it changes to 62 volts?
Have you measured the real battery voltage AT THE BATTERY terminals, NOT the Classic.
I just cannot believe the battery suddenly changes from 55.8 volts to 62 volts when the power input from the Classic changes from 45 watts to 90 watts.

Have you tried another voltmeter?  So that we can eliminate a faulty voltmeter.

dgd

[boB]

Very interesting.  Thank you Dmitry for the video.  This is usually a great way to show problems.

Is it possible that your meter is set to peak-hold mode ??   I am wondering if maybe a
load came on at 6:19 where I briefly see the MNGP show "Bulk MPPT" with the Classic
Vbatt meter dropping from 56.2V down to something in the 55.x volt range.

With a small battery size (14 amp-hour) the Classic might possibly overshoot and
show a peak voltage of around 62 volts but you would not necessarily see this
show up on the Classic's average readout of Vbatt because it is so quick.
Also, I notice that your meter shows a high voltage after you remove the probes
from the Classic's blue terminal block.

Interesting problem, especially if your meter is not set for peak (maximum) hold.
I will wait for your reply about possible peak reading mode.

[Dmitry]

boB thanks.
Once again began to review the video and just now noticed.
At the time 6:19 of the voltage controller briefly changed the type of work c "Аbsorb" on "Bulk MPPT" then again "Аbsorb".

About your comments.
I can safely say that correctly the voltmeter to measure the voltage, and is not installed in the remembering of the peak voltage.
Maybe some error in measurement, but it is on my memory was not more of 0.4 volts.
Look closely video, 8:40. I specifically demonstrated test measurements, disconnecting and connecting a voltmeter to the terminals of the controller.
Note 9:08 I measured the voltage at the terminals coming from solar panels, the difference amounted to 0.7 volts.
Then again I measured the voltage at the terminals of the battery and the difference in 6 volts present.
About the high voltage at the end of the video of which you speak.
Voltmeter is in automatic mode voltage selection and tried to measure chaotic induced voltage test leads, not connected to anything. This is not volts you only saw the mV, I write 10:05 specially shorted out probes voltmeter and indications steel 0, had this zapechatlet on the video, I at that time about it I just wasn't thinking.

Don't get me wrong.
To me there is no benefit to fool your head and try to falsify the results on the video.
Just now, the controller works, I think, wrong, and I fear to leave it unattended. He increased voltage burn either batteries or himself.

Ready to any experiments, describe what and how to spend, I will try to do with the video report. If you want, I'm ready to record video with two testers different manufacturers.

As you know I'm not very English, so if you will write something difficult to understand, try to write simpler.

[boB]

I know you are not wanting to fool me on any of this but I must ask questions just the same so that
I fully understand what is going on.

What I do not understand then is why your meter is reading 62 volts when the Classic is reading
56.2 volts.  I could understand if the Classic was reading 62 volts and your meter was reading 56.2
volts though, maybe.

Maybe you can wiggle the blue terminal block and see i the readings are different when
it is moved.

Also, can you please take some insulated tool or piece of plastic and lightly tap the Classic's
circuit board to see if it has some kind of loose connection or loose component ?
I would try tapping lightly between the lower left and over to the large 100 pin
microprocessor to see if the Classic's input voltage changes at all.

Does your system have another charger on it besides the Classic during these tests ?

[Dmitry]

boB understand your doubts.
Watch a video from 9:00 I specifically measured the voltage on the blue terminal from solar panels, what would you have had the impression that my voltmeter measures the wrong voltage.
At 9:10 input voltage which shows panel controller 114,8 volts, my tester shows -115,2 volts.
Then at 9:18 I change the polarity measurements rearranging leads tester, tension controller to 114.7 Volta, my voltmeter 115,3 volts.
Notice the difference in absolute value in the input voltage of the solar panels in the first case amounted to 0.4 volts, the second of 0.6 volts.
Next, measure the voltage at the battery terminals.
At 9:29 have the voltage of the controller shows 56,2 volts measured by the tester -62,1 volts.
At 9:37 have the voltage of the controller shows 56,2 volts measured by the tester 62,1 volts.
Total difference in absolute value stresses amounted to 5.9 volts.
Even with the maximum error that was me recorded in the measurement of voltages of solar batteries of 0.6 volts, the difference 5.8 volts cannot be an error of the tester.
All miralos direct touch the terminals to exclude any assumptions about the voltage drop on the cable.

As I wrote previously, the video was shown easy charging of batteries.
Any stress, extra charger not connected.
Battery capacity video 14 ampere hour 48 volt , was chosen exclusively for quick demonstration of the process.

I feel blue terminal strip soldered so thoroughly that even move would be difficult. I suspect that maybe the controller somewhere bad contact, but in appearance are not.

There is a big question on black strap which screwed power keys.  I controller came and she in several places between the screw is visible cracks. But all tightly curled. On this occasion I did not write and do not turn, during the unpacking just did a photo where it is clearly visible.
All the rest are collected just fine, nothing to loose, all filled with varnish.
The controller is so advanced, there are so many different elements that go into it, I think it is simply harmful.

I went and still do not have direct access to the controller. He turned off. My relatives who he is, do not have sufficient skills to work with such sophisticated equipment, so I have forbidden anything to do with it.

Write all what you think is necessary, I am ready as soon as you arrive and I will have access to the controller, perform any measurement of change of the mode of work, even if I need to make updates to the software.

[boB]

The cracks on the black FET hold down clamp (strip) is normal and OK.  There is metal inside of it.


There is another way to see what the Classic thinks the battery voltage is in addition to the upper right
Vbatt reading when the Classic is ON.  When I say "ON", I mean that it is displaying MPPT or Absorb
and not Resting.

From the main status screen (you were showing this in your video)  press and hold down the Left-Arrow key...
Then, tap (press) the ENTER key while holding down the left-arrow key.  This will bring up another screen
with numbers on it.  It will show IN on the top left and OUT on the top right and a + and - in the middle
with numbers to the left and right of the + and -      Look at the lower right number in the lower right
corner.  This will be the battery voltage without the decimal point.  (a ","  in your language of course)...

So, I would like to know what that lower right number shows when you see the 62.1 volts on your meter.
The lower right number in this test screen is read differently than the main battery voltage that you
were showing in your video.  Let's try to see if this lower-right number is different than the
main status screen battery voltage number when your meter shows 62 volts as well as when
the Classic battery voltage is showing 55 or 56 volts.

That new lower right number would show 621 if it thinks that the battery is  62.1  or  62,1  volts
and  562 if it thinks that the battery voltage is 56,2 volts.   It will most likely be slightly different
than the main battery voltage but maybe it will show near 62 volts when your meter is showing
62 volts.

Please give this a try.